The garment industry is responsible for approximately 4% of global emissions, mostly from the production process and is extremely resource-intensive (1). In response, a system of voluntary-sustainability standards (VSS) has emerged. Companies pledge to meet certain environmental standards, such as specific standards for sourcing organic cotton. However, these third-party standards are actors in their own right - often managed through influential non-profits and international organizations. Dynamics between a standard-setter and an apparel company have not yet been broadly explored. This blog post introduces a proof of concept for the quantitative aspect of my master’s thesis that will explore the dynamics between VSS and garment companies.
The network that I am studying is a two-mode, bipartite network where companies and VSS are distinct node sets and edges form between the two, but not within the groups. For the context of this class assignment, I needed to limit my data collection so I decided to limit the node set by company, selecting a subset of 11 apparel companies and then developing an edgelist based on VSS that they have adopted. I based my subset on the McKinsey Global Fashion Index (MGFI), a list of the top apparel companies based on average economic profit covering 2019 and 2020 (2).
I want to note that there is correlation between the top companies by revenue versus profit, but the two lists aren’t identical so my list excludes actors like H&M that did not rank on MGFI. Furthermore, MGFI looks at publicly-held companies because information is more readily available, compared to the relative opacity of privately-held companies such as Shein (3). However, these “ultra-fast-fashion” companies are quickly growing in popularity. Future iterations of my dataset will need to account for both revenue and profit - potentially comparing if profitability makes a difference in what standards are adopted and for privately-held brands with limited publicly-accessible data.
Once I selected my 11 companies, I then went through all of their most recent ESG reports, annual reports, and press releases and developed my edgelist based on which VSS they publicly disclosed adopting (4). Part of my broader hypothesis is that brands adopt VSS as a result of consumer and shareholder pressure, thus motivating the public disclosure of VSS adoption, facilitating my data gathering. I also verified on the websites of VSS standards to ensure the most accurate data collection possible (5). Obviously, this form of data gathering is prone to mistakes. For instance, as I am interested in environmental standards, I excluded labor standards, but there is a possibility that certain labor standards also have environmental components.
The types of products sold also impacts VSS adoption and this is not accounted for in my dataset. A company like Lululemon that does not sell products incorporating gold will naturally not adopt gold VSS.
At the end of my data gathering, I had built a bipartite network with 11 company nodes with the nodal attributes of
2019/2020 profitability in USD
luxury company (y/n), stock-index
country of legal domicile
Sustainalytics ranking, an sustainability scoring index (9).
The VSS node set includes 66 distinct standards with three nodal attributes.
VSS-area (“Materials”, “Water”, “Reporting”, “Chemistry”, “Emissions”, “Energy”,“General”, “Waste”, “Forestry”)
country of legal domicile
listed within Textile Exchange.
There are 181 edges between the two node sets.
As we can see from the first visualization, Kering looks like it is the company with the highest degree of VSS standards. Furthermore, the majority of VSS are materials related.
Looking at degrees, Kering is confirmed to be the company with the highest number of out-ties, 44. UN Global Compact, FSC, ZDHC, Fashion Industry Charter, and Leather Working Group share the most number of in-ties for VSS, 7.